{
    "object_id": "https://biocomputeobject.org/GLY_000084/v-1.12.3",
    "spec_version": "1.3.0",
    "extension_domain": [
        {
            "extension_schema": "http://www.w3id.org/biocompute/extension_domain/1.1.0/dataset/dataset_extension.json",
            "dataset_extension": {
                "additional_license": {
                    "data_license": "https://creativecommons.org/licenses/by/4.0/",
                    "script_license": "https://www.gnu.org/licenses/gpl-3.0.en.html"
                },
                "dataset_categories": [
                    {
                        "category_value": "Homo sapiens",
                        "category_name": "species"
                    },
                    {
                        "category_value": "Protein",
                        "category_name": "molecule"
                    },
                    {
                        "category_value": "Diseases_UniProtKB",
                        "category_name": "tags"
                    },
                    {
                        "category_value": "csv",
                        "category_name": "file_type"
                    },
                    {
                        "category_value": "Reviewed",
                        "category_name": "status"
                    },
                    {
                        "category_value": "internal",
                        "category_name": "scope"
                    }
                ]
            }
        }
    ],
    "provenance_domain": {
        "license": "https://creativecommons.org/licenses/by/4.0/",
        "name": "Human Protein Diseases (UniProtKB)",
        "contributors": [
            {
                "affiliation": "The George Washington University",
                "contribution": [
                    "createdBy"
                ],
                "name": "Jeet Vora",
                "email": "jeetvora@gwu.edu"
            },
            {
                "affiliation": "The George Washington University",
                "contribution": [
                    "createdBy"
                ],
                "name": "Rahi Navelkar",
                "email": "rsn13@gwu.edu"
            },
            {
                "affiliation": "The George Washington University",
                "contribution": [
                    "createdBy"
                ],
                "name": "Robel Kahsay",
                "email": "rykahsay@gwu.edu"
            },
            {
                "affiliation": "The George Washington University",
                "contribution": [
                    "createdBy"
                ],
                "name": "Karina Martinez",
                "email": "karinamartinez@gwu.edu"
            },
            {
                "affiliation": "The George Washington University",
                "contribution": [
                    "createdBy"
                ],
                "name": "Amanda Bell",
                "email": "amandab2140@gwu.edu"
            },
            {
                "affiliation": "The George Washington University",
                "contribution": [
                    "createdBy"
                ],
                "name": "Jonathon Keeney",
                "email": "keeneyjg@gwu.edu"
            },
            {
                "affiliation": "The George Washington University",
                "contribution": [
                    "createdBy"
                ],
                "name": "Millicent Quartey",
                "email": "mquartey@email.gwu.edu"
            },
            {
                "affiliation": "",
                "contribution": [
                    "contributedBy"
                ],
                "name": "Not Applicable",
                "email": ""
            }
        ],
        "created": "2019-01-23T17:21:05.953432",
        "modified": "2022-01-12T16:04:35.397885",
        "version": "v-1.12.3"
    },
    "description_domain": {
        "keywords": [
            "disease",
            "HPO",
            "OMIM",
            "Genomics England PanelApp",
            "Monarch Initiative",
            "MONDO",
            "DO"
        ],
        "pipeline_steps": [
            {
                "name": "make-protein-dataset.py",
                "step_number": 1,
                "version": "",
                "input_list": [
                    {
                        "uri": "https://data.glygen.org/ln2data/releases/data/v-1.12.3/reviewed/human_protein_masterlist.csv",
                        "filename": "human_protein_masterlist.csv"
                    },
                    {
                        "uri": "https://data.glygen.org/ln2data/releases/data/v-1.12.3/source/uniprot-proteome-homo-sapiens.nt",
                        "filename": "uniprot-proteome-homo-sapiens.nt"
                    },
                    {
                        "uri": "https://data.glygen.org/ln2data/releases/data/v-1.12.3/reviewed/protein_domap.csv",
                        "filename": "protein_domap.csv"
                    }
                ],
                "output_list": [
                    {
                        "uri": "https://data.glygen.org/ln2data/releases/data/v-1.12.3/reviewed/human_protein_disease_uniprotkb.csv",
                        "filename": "human_protein_disease_uniprotkb.csv"
                    }
                ],
                "prerequisite": [],
                "description": "The script extracts the OMIM phenotypes for UniProtKB accessions and MONDO disease information from human_glyco-phenotype_disease_curation_list_v12-18-74.csv and Genomics England disease data from . The script maps the UniProtKB accessions to the UniprotKB canonical accessions."
            }
        ]
    },
    "execution_domain": {
        "external_data_endpoints": [],
        "environment_variables": {},
        "script_driver": "manual",
        "software_prerequisites": [
            {
                "version": "2.7.5",
                "name": "Python",
                "uri": {
                    "uri": "https://www.python.org/download/releases/2.7.5/"
                }
            }
        ],
        "script": [
            {
                "uri": {
                    "uri": "https://github.com/GW-HIVE/glygen-backend-integration/blob/master/pipeline/integrator/make-protein-dataset.py",
                    "filename": "make-protein-dataset.py"
                }
            }
        ]
    },
    "error_domain": {
        "empirical_error": {},
        "algorithmic_error": {},
        "emperical_error": {}
    },
    "parametric_domain": [
        {
            "step": "",
            "param": "",
            "value": ""
        }
    ],
    "usability_domain": [
        "The Human Protein Diseases dataset contains list of human [taxid:9606] diseases from OMIM, MONDO and Genomics England databases mapped to the UniProtKB protein canonical accessions and Disease Ontology (DO) identifiers..",
        "Online Mendelian Inheritance in Man (OMIM) is a continuously updated catalog of human genes and genetic disorders and traits, with a particular focus on the gene-phenotype relationship.",
        "The Monarch Initiative is focused primarily on phenotype-related resources and brings in data associated with those phenotypes so that the users can begin to make connections among other biological entities of interest.",
        "The Disease Ontology has been developed as a standardized ontology for human disease with the purpose of providing the biomedical community with consistent, reusable and sustainable descriptions of human disease terms, phenotype characteristics and related medical vocabulary disease concepts through collaborative efforts of biomedical researchers, coordinated by the University of Maryland School of Medicine, Institute for Genome Sciences.",
        "If you are using this dataset please give proper attribution to OMIM, Monarch Initiative, DO, and EMBL-EBI-UniProt and GlyGen"
    ],
    "etag": "f9d93b4f6c396d52439a0ffbaebb76f4",
    "io_domain": {
        "input_subdomain": [
            {
                "uri": {
                    "filename": "protein_disease_names.csv",
                    "uri": "https:/data.glygen.org/ln2data/releases/data/v-1.12.3/reviewed/protein_disease_names.csv"
                }
            },
            {
                "uri": {
                    "filename": "protein_disease_idmap.csv",
                    "uri": "https:/data.glygen.org/ln2data/releases/data/v-1.12.3/reviewed/protein_disease_idmap.csv"
                }
            },
            {
                "uri": {
                    "filename": "human_glyco-phenotype_disease_curation_list_v12-18-74.csv",
                    "uri": "https:/data.glygen.org/ln2downloads/ohsu/human_glyco-phenotype_disease_curation_list_v12-18-74.csv"
                }
            },
            {
                "uri": {
                    "filename": "uniprot-proteome-homo-sapiens.nt",
                    "uri": "https:/data.glygen.org/ln2downloads/ebi/2021_04/uniprot-proteome-homo-sapiens.nt"
                }
            }
        ],
        "output_subdomain": [
            {
                "mediatype": "csv",
                "uri": {
                    "uri": "https:/data.glygen.org/ln2data/releases/data/v-1.12.3/reviewed/human_protein_disease_uniprotkb.csv",
                    "filename": "human_protein_disease_uniprotkb.csv"
                }
            },
            {
                "mediatype": "csv",
                "uri": {
                    "uri": "https:/data.glygen.org/ln2data/releases/data/v-1.12.3/reviewed/human_protein_disease_uniprotkb.stat.csv",
                    "filename": "human_protein_disease_uniprotkb.stat.csv"
                }
            }
        ]
    }
}